Hacker News Reader

Comment by puzzledobserver | Hacker News Reader

Is ranking by brand necessarily the most useful ranking?

How does the Toyota Prius (with 19 potential trouble areas, per the article) compare with a non-hybrid (say a Mazda 3)?

The first graph indicates that Toyotas are, on the whole, more reliable than Mazdas, and the second graph indicates that hybrid drivetrains (despite having more trouble points) are marginally more reliable than purely gas powered vehicles.

Taken together, this suggests that a recent Prius would be more reliable than a recent Mazda 3, but it would be nice to have a model-wise reliability score, no?

View on HN

Here is a similar ranking from Germany where you can sort by class of car and than compare car models instead of manufacturers (just scroll down a bit to get to that part and have it translated by your browser): https://www.adac.de/rund-ums-fahrzeug/unfall-schaden-panne/a...

This is really good. "Number of breakdowns per 1,000 registered vehicles per first registration year:"

Funny how the order is almost the opposite of the rather opaque consumerreports one. While looking at actual breakdowns is more telling of major issues, looking at only the first year isn't going to reveal much beyond glaring manufacturing defects. I'd be interested in more long term issues (although understandably this would be more of a past performance based on previous models metric). Although long term all cars reliability (especially ICE) is an AND operation with maintenance and care, but with a large enough sample size there's a hope this noise should average out.. ignoring the annoying issue of buyer bias.

The article also points out the vast majority of breakdowns are from dead batteries - Whether that's user error or parasitical leaks, or bad batteries is not clear, but sticking in a bigger battery with high cold cranking amps is a very cheap fix for that. Parasitical leaks are more of a pain to solve, especially since fitting a kill switch tends to make the computers in modern cars very upset.

Not sure if you noticed, but you can see the issues for each car by the year it was first registered in, so if a model is out for a few years you will be able to say more about long term reliability.

If I had to guess it is partially due to user error, but the majority probably stems from people not replacing them when they should and being one of the few things in newer cars where you need to actually call help and can’t wait a few days to have it repaired.

Reply on news.ycombinator.com